How Effect.AI can outperform the top AI innovators
Who develops AI currently?
Effect.AI - How blockchain technology can outperform current leaders in AI
How Effect.AI can outperform the top AI innovators
AIBrain.jpg
Image Source: Pixabay
Who develops AI currently?
AI has begun to exist in almost every aspect of everyday life. From Amazon's algorithms for suggested products you might also like, to university research projects, to an individual trying to build a trading bot that combines the predictability of patterns in technical analysis with the disruptive nature of human response to positive/negative news stories, AI is programmed for nearly anything quantifiable. Although the development of AI is wide-reaching across many industries, the most significant and complex AI developments that will shape the future are mostly done by a much more restrictive group. This was noted by a new blockchain company called Effect.AI who had this to say in their lightpaper:
"However, while some academic achievements (with limited use) in the field of AI are available to the public, the majority of AI development happens behind closed doors at large corporations." - Effect.AI lightpaper
So, whereas AI is used all over the place, the most elaborate and significant new developments are not developed in the public eye, but by corporations. I'm talking about entities such as Google, Amazon, IBM, and others. One example is the IBM's Watson AI that has recently been advertised on TV. Although this project has been running for quite some time, they're now starting to bring it forward into the public's eye. Check out this ad they ran in 2017:
Source: IBM UK-EN
If you prefer a real world application, check out this classic Watson experiment competing in Jeopardy:
Effect.AI - How blockchain technology can outperform current leaders in AI
How Effect.AI can outperform the top AI innovators
AIBrain.jpg
Image Source: Pixabay
Who develops AI currently?
AI has begun to exist in almost every aspect of everyday life. From Amazon's algorithms for suggested products you might also like, to university research projects, to an individual trying to build a trading bot that combines the predictability of patterns in technical analysis with the disruptive nature of human response to positive/negative news stories, AI is programmed for nearly anything quantifiable. Although the development of AI is wide-reaching across many industries, the most significant and complex AI developments that will shape the future are mostly done by a much more restrictive group. This was noted by a new blockchain company called Effect.AI who had this to say in their lightpaper:
"However, while some academic achievements (with limited use) in the field of AI are available to the public, the majority of AI development happens behind closed doors at large corporations." - Effect.AI lightpaper
So, whereas AI is used all over the place, the most elaborate and significant new developments are not developed in the public eye, but by corporations. I'm talking about entities such as Google, Amazon, IBM, and others. One example is the IBM's Watson AI that has recently been advertised on TV. Although this project has been running for quite some time, they're now starting to bring it forward into the public's eye. Check out this ad they ran in 2017:
Source: IBM UK-EN
If you prefer a real world application, check out this classic Watson experiment competing in Jeopardy:
Source: IBM - Watson
Similar projects can be found within the other large organizations I mentioned as well. Some are more well known than others, such as the Google Assistant or Amazon's Alexa. People all over the world are using products that were born from AI research. But you'll notice that the most popular AI driven applications being put into use were all developed by these big corporations. Is it possible that these corporations just happen to have all the best talent and should therefore reap all of the rewards of this development? The answer is no, and I can prove it!
Netflix, Ensemble Effect, and a Book Recommendation
In Eric Siegel's 2013 book, "Predictive Analytics: the power to predict who will click, buy, lie, or die," he has an entire chapter dedicated to the results of crowdsourcing and the ensemble effect. In this chapter he gives a real world example of a 2006 contest sponsored by Netflix with the intent to optimize their movie recommendations. They offered a $1M prize to whatever team could improve their current algorithm by 10%. I highly recommend picking this book up, it's a very enlightening and easy to read gold mine of information in the field of predictive analytics.
The contest that is the primary topic (among some other similar examples) in this chapter talks about how each of the teams that entered the competition were struggling to make any significant improvements on their original models as there was a point of diminishing returns on extra effort within the same model. That was, until a couple of teams who had hit those diminishing returns decided to link up and combine their predictive models into one, launching them into the front of the pack and closing in on the 10% goal. Eric writes:
"For the Netflix Prize, the rules of the game had now changed, triggering a new flurry of merging and blending as teams consolidated, rolling up into bigger and better competitors." - Eric Siegel (Predictive Analytics: the power to predict who will click, buy, lie, or die)
That same team ended up absorbing yet another high end competitor to be the first to complete the challenge and ultimately split the $1M prize. He goes on to explain this as the "Ensemble Effect" and discloses how competitors can turn into collaborators by forming one model that incorporates 2 or more already somewhat successful models. He also shows how these combined models, once tweaked, will most often outperform each of the individual modules.
How Effect.AI will outperform competitors
Effect.AI - How blockchain technology can outperform current leaders in AI
mjo (42) in cryptocurrency • 2 months ago
Logo.JPG
How Effect.AI can outperform the top AI innovators
AIBrain.jpg
Image Source: Pixabay
Who develops AI currently?
AI has begun to exist in almost every aspect of everyday life. From Amazon's algorithms for suggested products you might also like, to university research projects, to an individual trying to build a trading bot that combines the predictability of patterns in technical analysis with the disruptive nature of human response to positive/negative news stories, AI is programmed for nearly anything quantifiable. Although the development of AI is wide-reaching across many industries, the most significant and complex AI developments that will shape the future are mostly done by a much more restrictive group. This was noted by a new blockchain company called Effect.AI who had this to say in their lightpaper:
"However, while some academic achievements (with limited use) in the field of AI are available to the public, the majority of AI development happens behind closed doors at large corporations." - Effect.AI lightpaper
So, whereas AI is used all over the place, the most elaborate and significant new developments are not developed in the public eye, but by corporations. I'm talking about entities such as Google, Amazon, IBM, and others. One example is the IBM's Watson AI that has recently been advertised on TV. Although this project has been running for quite some time, they're now starting to bring it forward into the public's eye. Check out this ad they ran in 2017:
Source: IBM UK-EN
If you prefer a real world application, check out this classic Watson experiment competing in Jeopardy:
Source: IBM - Watson
Similar projects can be found within the other large organizations I mentioned as well. Some are more well known than others, such as the Google Assistant or Amazon's Alexa. People all over the world are using products that were born from AI research. But you'll notice that the most popular AI driven applications being put into use were all developed by these big corporations. Is it possible that these corporations just happen to have all the best talent and should therefore reap all of the rewards of this development? The answer is no, and I can prove it!
Netflix, Ensemble Effect, and a Book Recommendation
In Eric Siegel's 2013 book, "Predictive Analytics: the power to predict who will click, buy, lie, or die," he has an entire chapter dedicated to the results of crowdsourcing and the ensemble effect. In this chapter he gives a real world example of a 2006 contest sponsored by Netflix with the intent to optimize their movie recommendations. They offered a $1M prize to whatever team could improve their current algorithm by 10%. I highly recommend picking this book up, it's a very enlightening and easy to read gold mine of information in the field of predictive analytics.
The contest that is the primary topic (among some other similar examples) in this chapter talks about how each of the teams that entered the competition were struggling to make any significant improvements on their original models as there was a point of diminishing returns on extra effort within the same model. That was, until a couple of teams who had hit those diminishing returns decided to link up and combine their predictive models into one, launching them into the front of the pack and closing in on the 10% goal. Eric writes:
"For the Netflix Prize, the rules of the game had now changed, triggering a new flurry of merging and blending as teams consolidated, rolling up into bigger and better competitors." - Eric Siegel (Predictive Analytics: the power to predict who will click, buy, lie, or die)
That same team ended up absorbing yet another high end competitor to be the first to complete the challenge and ultimately split the $1M prize. He goes on to explain this as the "Ensemble Effect" and discloses how competitors can turn into collaborators by forming one model that incorporates 2 or more already somewhat successful models. He also shows how these combined models, once tweaked, will most often outperform each of the individual modules.
How Effect.AI will outperform competitors
It's not for the sake of promoting a good book that I bring up the previous Netflix example. There are a couple major reasons why it's relevant here. First, the crowdsourcing option went on to prove that opening an issue to the public will produce better results than keeping these developments within the organization. Right off the bat, there were individuals/groups that were able to beat the Netflix algorithm (albeit not by 10% yet) to show that within the confines of your organization you likely don't have the best answer to your most complex problems. Effect.AI will capitalize on this same concept, having the ability for crowdsourcing the ability to "teach" the AI.
Image Source: Pixabay
Another major reason Effect.AI stands to topple the industry giants is the extension of the crowdsourcing benefit which is the collaborative potential among top contributors, referred to by Eric Siegel in his book as the ensemble effect. Let's say that Google and a company using Effect.AI both have the same extraordinarily difficult task to perform. Google certainly has a large, well funded, and well educated group to perform this task. They will likely start off with a bang and hit close to the target within a short period of time. However, much like in the Netflix study above they'll likely begin to see diminishing returns on additional effort in their current model. Perhaps it takes them years of research and development to ultimately achieve their goal. Meanwhile, on Effect.AI a couple hundred competent developers give the task a shot to find that they fall short of the goal at first as well. Perhaps they all try to work out the major kinks in their projects until progress slows down. Then, unlike Google, they can reach out to others tackling the same task to discuss partnerships. Perhaps instead of years of inching to the finish line that Google is likely to face, the groups within the Effect.AI platform are able to complete the task within 1 year and move onto the next task due to their ability to collaborate.
How does Effect.AI work?
Effect.AI is a blockchain solution for AI development and usage. It creates a marketplace where users can contribute to the development of AI solutions as well as a medium for leveraging such AI solutions for real world applications. It is built using NEO smart contracts. The Effect.AI solution will be rolled out in 3 phases.
Source: Effect.AI website
For more information on these phases, click here. Also, here's an explanitory video from Effect.AI.
Effect.AI - How blockchain technology can outperform current leaders in AI
mjo (42) in cryptocurrency • 2 months ago
Logo.JPG
How Effect.AI can outperform the top AI innovators
AIBrain.jpg
Image Source: Pixabay
Who develops AI currently?
AI has begun to exist in almost every aspect of everyday life. From Amazon's algorithms for suggested products you might also like, to university research projects, to an individual trying to build a trading bot that combines the predictability of patterns in technical analysis with the disruptive nature of human response to positive/negative news stories, AI is programmed for nearly anything quantifiable. Although the development of AI is wide-reaching across many industries, the most significant and complex AI developments that will shape the future are mostly done by a much more restrictive group. This was noted by a new blockchain company called Effect.AI who had this to say in their lightpaper:
"However, while some academic achievements (with limited use) in the field of AI are available to the public, the majority of AI development happens behind closed doors at large corporations." - Effect.AI lightpaper
So, whereas AI is used all over the place, the most elaborate and significant new developments are not developed in the public eye, but by corporations. I'm talking about entities such as Google, Amazon, IBM, and others. One example is the IBM's Watson AI that has recently been advertised on TV. Although this project has been running for quite some time, they're now starting to bring it forward into the public's eye. Check out this ad they ran in 2017:
Source: IBM UK-EN
If you prefer a real world application, check out this classic Watson experiment competing in Jeopardy:
Source: IBM - Watson
Similar projects can be found within the other large organizations I mentioned as well. Some are more well known than others, such as the Google Assistant or Amazon's Alexa. People all over the world are using products that were born from AI research. But you'll notice that the most popular AI driven applications being put into use were all developed by these big corporations. Is it possible that these corporations just happen to have all the best talent and should therefore reap all of the rewards of this development? The answer is no, and I can prove it!
Netflix, Ensemble Effect, and a Book Recommendation
In Eric Siegel's 2013 book, "Predictive Analytics: the power to predict who will click, buy, lie, or die," he has an entire chapter dedicated to the results of crowdsourcing and the ensemble effect. In this chapter he gives a real world example of a 2006 contest sponsored by Netflix with the intent to optimize their movie recommendations. They offered a $1M prize to whatever team could improve their current algorithm by 10%. I highly recommend picking this book up, it's a very enlightening and easy to read gold mine of information in the field of predictive analytics.
The contest that is the primary topic (among some other similar examples) in this chapter talks about how each of the teams that entered the competition were struggling to make any significant improvements on their original models as there was a point of diminishing returns on extra effort within the same model. That was, until a couple of teams who had hit those diminishing returns decided to link up and combine their predictive models into one, launching them into the front of the pack and closing in on the 10% goal. Eric writes:
"For the Netflix Prize, the rules of the game had now changed, triggering a new flurry of merging and blending as teams consolidated, rolling up into bigger and better competitors." - Eric Siegel (Predictive Analytics: the power to predict who will click, buy, lie, or die)
That same team ended up absorbing yet another high end competitor to be the first to complete the challenge and ultimately split the $1M prize. He goes on to explain this as the "Ensemble Effect" and discloses how competitors can turn into collaborators by forming one model that incorporates 2 or more already somewhat successful models. He also shows how these combined models, once tweaked, will most often outperform each of the individual modules.
How Effect.AI will outperform competitors
It's not for the sake of promoting a good book that I bring up the previous Netflix example. There are a couple major reasons why it's relevant here. First, the crowdsourcing option went on to prove that opening an issue to the public will produce better results than keeping these developments within the organization. Right off the bat, there were individuals/groups that were able to beat the Netflix algorithm (albeit not by 10% yet) to show that within the confines of your organization you likely don't have the best answer to your most complex problems. Effect.AI will capitalize on this same concept, having the ability for crowdsourcing the ability to "teach" the AI.
Group.jpg
Image Source: Pixabay
Another major reason Effect.AI stands to topple the industry giants is the extension of the crowdsourcing benefit which is the collaborative potential among top contributors, referred to by Eric Siegel in his book as the ensemble effect. Let's say that Google and a company using Effect.AI both have the same extraordinarily difficult task to perform. Google certainly has a large, well funded, and well educated group to perform this task. They will likely start off with a bang and hit close to the target within a short period of time. However, much like in the Netflix study above they'll likely begin to see diminishing returns on additional effort in their current model. Perhaps it takes them years of research and development to ultimately achieve their goal. Meanwhile, on Effect.AI a couple hundred competent developers give the task a shot to find that they fall short of the goal at first as well. Perhaps they all try to work out the major kinks in their projects until progress slows down. Then, unlike Google, they can reach out to others tackling the same task to discuss partnerships. Perhaps instead of years of inching to the finish line that Google is likely to face, the groups within the Effect.AI platform are able to complete the task within 1 year and move onto the next task due to their ability to collaborate.
How does Effect.AI work?
Effect.AI is a blockchain solution for AI development and usage. It creates a marketplace where users can contribute to the development of AI solutions as well as a medium for leveraging such AI solutions for real world applications. It is built using NEO smart contracts. The Effect.AI solution will be rolled out in 3 phases.
3Phases.JPG
Source: Effect.AI website
For more information on these phases, click here. Also, here's an explanitory video from Effect.AI.
Source: Effect.AI website
The possibilities are all but endless when it comes to AI development. Here's a few tasks that are highlighted on their website: image classification, audio transcription, and data labeling. Since the platform can be used both by task requesters and workers/users, there's plenty of potential for interesting and valuable work. When we're working with intellectual property, a blockchain based solution is very valuable because it leaves behind it an immutable record of contributions. It's this technology that makes it possible for Effect.AI to safetly bring these projects to the public.
Summary
These giant corporations that currently have a firm grasp on the AI space are likely filled with high end talent, which makes them intimidating and nearly impossible to defeat for individuals and small companies. Effect.AI has the advantage of being able to present a task to any of the 7 billion people on earth that have the capacity to perform the task. Then, the added ability to have a community that can collaborate for any tasks that are still just outside the grasp of the individuals/groups that originally take on the tasks. Let's say for example that Google vigorously hires one in a million talent for each position within the United States, that means there are nearly 325 (based on nearly 326 Million population) other equally qualified candidates not employed by Google, and that's looking at just the United States, scale that out to the whole globe and you'll see the potential pool of candidates is quite vast. If you put things in context, it's not really possible for any corporation to outsmart the pool that is the public at large. This is how the execution of the Effect.AI roadmap will position them for successes that are unachievable within the corporate AI paradigm we currently are restricted by.
Click here to view the 
effect.ai2018